Expansion of the Human Phenotype Ontology (HPO) knowledge base and resources

BDV Syndrome: An Emerging Syndrome With Profound Obesity and Neurodevelopmental Delay Resembling Prader-Willi Syndrome

CONTEXT

CPE encodes carboxypeptidase E, an enzyme that converts proneuropeptides and propeptide hormones to bioactive forms. It is widely expressed in the endocrine and central nervous system. To date, 4 individuals from 2 families with core clinical features including morbid obesity, neurodevelopmental delay, and hypogonadotropic hypogonadism, harboring biallelic loss-of-function (LoF) CPE variants, have been reported.

OBJECTIVE

We describe 4 affected individuals from 3 unrelated consanguineous families, 2 siblings of Syrian, 1 of Egyptian, and 1 of Pakistani descent, all harboring novel homozygous CPE LoF variants.

METHODS

After excluding Prader-Willi syndrome (PWS), exome sequencing was performed in both Syrian siblings. The variants identified in the other 2 individuals were reported as research variants in a large-scale exome study and in the ClinVar database. Computational modeling of all possible missense alterations allowed assessing CPE tolerance to missense variants.

RESULTS

All affected individuals were severely obese with neurodevelopmental delay and other endocrine anomalies. Three individuals from 2 families shared the same CPE homozygous truncating variant c.361C > T, p.(Arg121*), while the fourth carried the c.994del, p.(Ser333Alafs*22) variant. Comparison of clinical features with previously described cases and standardization according to the Human Phenotype Ontology terms indicated a recognizable clinical phenotype, which we termed Blakemore-Durmaz-Vasileiou (BDV) syndrome. Computational analysis indicated high conservation of CPE domains and intolerance to missense changes.

CONCLUSION

Biallelic truncating CPE variants are associated with BDV syndrome, a clinically recognizable monogenic recessive syndrome with childhood-onset obesity, neurodevelopmental delay, hypogonadotropic hypogonadism, and hypothyroidism. BDV syndrome resembles PWS. Our findings suggest missense variants may also be clinically relevant.

The importance of automation in genetic diagnosis: Lessons from analyzing an inherited retinal degeneration cohort with the Mendelian Analysis Toolkit (MATK)

PURPOSE

In Mendelian disease diagnosis, variant analysis is a repetitive, error-prone, and time consuming process. To address this, we have developed the Mendelian Analysis Toolkit (MATK), a configurable, automated variant ranking program.

METHODS

MATK aggregates variant information from multiple annotation sources and uses expert-designed rules with parameterized weights to produce a ranked list of potentially causal solutions. MATK performance was measured by a comparison between MATK-aided and human-domain expert analyses of 1060 families with inherited retinal degeneration (IRD), analyzed using an IRD-specific gene panel (589 individuals) and exome sequencing (471 families).

RESULTS

When comparing MATK-assisted analysis with expert curation in both the IRD-specific gene panel and exome sequencing (1060 subjects), 97.3% of potential solutions found by experts were also identified by the MATK-assisted analysis (541 solutions identified with MATK of 556 solutions found by conventional analysis). Furthermore, MATK-assisted analysis identified 114 additional potential solutions from the 504 cases unsolved by conventional analysis.

CONCLUSION

MATK expedites the process of identification of likely solving variants in Mendelian traits, and reduces variability stemming from human error and researcher bias. MATK facilitates data reanalysis to keep up with the constantly improving annotation sources and next-generation sequencing processing pipelines. The software is open source and available at https://gitlab.com/matthew_maher/mendelanalysis.

Artificial intelligence (AI)-assisted exome reanalysis greatly aids in the identification of new positive cases and reduces analysis time in a clinical diagnostic laboratory

PURPOSE

Artificial intelligence (AI) and variant prioritization tools for genomic variant analysis are being rapidly developed for use in clinical diagnostic testing. However, their clinical utility and reliability are currently limited. Therefore, we performed a validation of a commercial AI tool (Moon) and a comprehensive reanalysis of previously collected clinical exome sequencing cases using an open-source variant prioritization tool (Exomiser) and the now-validated AI tool to test their feasibility in clinical diagnostics.

METHODS

A validation study of Moon was performed with 29 positive cases determined by previous manual analysis. After validation, reanalysis was performed on 80 previously manually analyzed nondiagnostic exome cases using Moon. Finally, a comparison between Moon and Exomiser was completed regarding their ability to identify previously completed positive cases and to identify new positive cases.

RESULTS

Moon correctly selected the causal variant(s) in 97% of manually analyzed positive cases and identified 7 new positive cases. Exomiser correctly identified the causal gene in 85% of positive cases and agreed with Moon by ranking the new gene in its top 10 list 43% of the time.

CONCLUSION

The use of AI in diagnostic laboratories greatly enhances exome sequencing analysis by reducing analysis time and increasing the diagnostic rate.

HFIP: an integrated multi-omics data and knowledge platform for the precision medicine of heart failure

As the terminal clinical phenotype of almost all types of cardiovascular diseases, heart failure (HF) is a complex and heterogeneous syndrome leading to considerable morbidity and mortality. Existing HF-related omics studies mainly focus on case/control comparisons, small cohorts of special subtypes, etc., and a large amount of multi-omics data and knowledge have been generated. However, it is difficult for researchers to obtain biological and clinical insights from these scattered data and knowledge. In this paper, we built the Heart Failure Integrated Platform (HFIP) for data exploration, fusion analysis and visualization by collecting and curating existing multi-omics data and knowledge from various public sources and also provided an auto-updating mechanism for future integration. The developed HFIP contained 253 datasets (7842 samples), multiple analysis flow, and 14 independent tools. In addition, based on the integration of existing databases and literature, a knowledge base for HF was constructed with a scoring system for evaluating the relationship between molecular signals and HF. The knowledge base includes 1956 genes and annotation information. The literature mining module was developed to assist the researcher to overview the hotspots and contexts in basic and clinical research. HFIP can be used as a data-driven and knowledge-guided platform for the basic and clinical research of HF.

Database URL: http://heartfailure.medical-bigdata.com

100,000 Genomes Pilot on Rare-Disease Diagnosis in Health Care - Preliminary Report

BACKGROUND

The U.K. 100,000 Genomes Project is in the process of investigating the role of genome sequencing in patients with undiagnosed rare diseases after usual care and the alignment of this research with health care implementation in the U.K. National Health Service. Other parts of this project focus on patients with cancer and infection.

METHODS

We conducted a pilot study involving 4660 participants from 2183 families, among whom 161 disorders covering a broad spectrum of rare diseases were present. We collected data on clinical features with the use of Human Phenotype Ontology terms, undertook genome sequencing, applied automated variant prioritization on the basis of applied virtual gene panels and phenotypes, and identified novel pathogenic variants through research analysis.

RESULTS

Diagnostic yields varied among family structures and were highest in family trios (both parents and a proband) and families with larger pedigrees. Diagnostic yields were much higher for disorders likely to have a monogenic cause (35%) than for disorders likely to have a complex cause (11%). Diagnostic yields for intellectual disability, hearing disorders, and vision disorders ranged from 40 to 55%. We made genetic diagnoses in 25% of the probands. A total of 14% of the diagnoses were made by means of the combination of research and automated approaches, which was critical for cases in which we found etiologic noncoding, structural, and mitochondrial genome variants and coding variants poorly covered by exome sequencing. Cohortwide burden testing across 57,000 genomes enabled the discovery of three new disease genes and 19 new associations. Of the genetic diagnoses that we made, 25% had immediate ramifications for clinical decision making for the patients or their relatives.

CONCLUSIONS

Our pilot study of genome sequencing in a national health care system showed an increase in diagnostic yield across a range of rare diseases. (Funded by the National Institute for Health Research and others.).

Phenotype-tissue expression and exploration (PTEE) resource facilitates the choice of tissue for RNA-seq-based clinical genetics studies

BACKGROUND

RNA-seq emerges as a valuable method for clinical genetics. The transcriptome is "dynamic" and tissue-specific, but typically the probed tissues to analyze (TA) are different from the tissue of interest (TI) based on pathophysiology.

RESULTS

We developed Phenotype-Tissue Expression and Exploration (PTEE), a tool to facilitate the decision about the most suitable TA for RNA-seq. We integrated phenotype-annotated genes, used 54 tissues from GTEx to perform correlation analyses and identify expressed genes and transcripts between TAs and TIs. We identified skeletal muscle as the most appropriate TA to inquire for cardiac arrhythmia genes and skin as a good proxy to study neurodevelopmental disorders. We also explored RNA-seq limitations and show that on-off switching of gene expression during ontogenesis or circadian rhythm can cause blind spots for RNA-seq-based analyses.

CONCLUSIONS

PTEE aids the identification of tissues suitable for RNA-seq for a given pathology to increase the success rate of diagnosis and gene discovery. PTEE is freely available at https://bioinf.eva.mpg.de/PTEE/.

Mouse Genome Informatics (MGI): latest news from MGD and GXD

The Mouse Genome Informatics (MGI) database system combines multiple expertly curated community data resources into a shared knowledge management ecosystem united by common metadata annotation standards. MGI's mission is to facilitate the use of the mouse as an experimental model for understanding the genetic and genomic basis of human health and disease. MGI is the authoritative source for mouse gene, allele, and strain nomenclature and is the primary source of mouse phenotype annotations, functional annotations, developmental gene expression information, and annotations of mouse models with human diseases. MGI maintains mouse anatomy and phenotype ontologies and contributes to the development of the Gene Ontology and Disease Ontology and uses these ontologies as standard terminologies for annotation. The Mouse Genome Database (MGD) and the Gene Expression Database (GXD) are MGI's two major knowledgebases. Here, we highlight some of the recent changes and enhancements to MGD and GXD that have been implemented in response to changing needs of the biomedical research community and to improve the efficiency of expert curation. MGI can be accessed freely at http://www.informatics.jax.org .

Prenatal phenotype of PNKP-related primary microcephaly associated with variants affecting both the FHA and phosphatase domain

Biallelic PNKP variants cause heterogeneous disorders ranging from neurodevelopmental disorder with microcephaly/seizures to adult-onset Charcot–Marie–Tooth disease. To date, only postnatal descriptions exist. We present the first prenatal diagnosis of PNKP-related primary microcephaly. Pathological examination of a male fetus in the 18th gestational week revealed micrencephaly with extracerebral malformations and thus presumed syndromic microcephaly. A recessive disorder was suspected because of previous pregnancy termination for similar abnormalities. Prenatal trio-exome sequencing identified compound heterozygosity for the PNKP variants c.498G>A, p.[(=),0?] and c.302C>T, p.(Pro101Leu). Segregation confirmed both variants in the sister fetus. Through RNA analyses, we characterized exon 4 skipping affecting the PNKP forkhead-associated (FHA) and phosphatase domains (p.Leu67_Lys166del) as the predominant effect of the paternal c.498G>A variant. We retrospectively investigated two unrelated individuals diagnosed with biallelic PNKP-variants to compare prenatal/postnatal phenotypes. Both carry the splice donor variant c.1029+2T>C intrans with a variant in the FHA domain (c.311T>C, p.(Leu104Pro); c.151G>C, p.(Val51Leu)). RNA-seq showed complex splicing for c.1029+2T>C and c.151G>C. Structural modeling revealed significant clustering of missense variants in the FHA domain with variants generating structural damage. Our clinical description extends the PNKP-continuum to the prenatal stage. Investigating possible PNKP-variant effects using RNA and structural modeling, we highlight the mutational complexity and exemplify a PNKP-variant characterization framework.

Deep semi-supervised learning ensemble framework for classifying co-mentions of human proteins and phenotypes

Background

Identifying human protein-phenotype relationships has attracted researchers in bioinformatics and biomedical natural language processing due to its importance in uncovering rare and complex diseases. Since experimental validation of protein-phenotype associations is prohibitive, automated tools capable of accurately extracting these associations from the biomedical text are in high demand. However, while the manual annotation of protein-phenotype co-mentions required for training such models is highly resource-consuming, extracting millions of unlabeled co-mentions is straightforward.

Results

In this study, we propose a novel deep semi-supervised ensemble framework that combines deep neural networks, semi-supervised, and ensemble learning for classifying human protein-phenotype co-mentions with the help of unlabeled data. This framework allows the ability to incorporate an extensive collection of unlabeled sentence-level co-mentions of human proteins and phenotypes with a small labeled dataset to enhance overall performance. We develop PPPredSS, a prototype of our proposed semi-supervised framework that combines sophisticated language models, convolutional networks, and recurrent networks. Our experimental results demonstrate that the proposed approach provides a new state-of-the-art performance in classifying human protein-phenotype co-mentions by outperforming other supervised and semi-supervised counterparts. Furthermore, we highlight the utility of PPPredSS in powering a curation assistant system through case studies involving a group of biologists.

Conclusions

This article presents a novel approach for human protein-phenotype co-mention classification based on deep, semi-supervised, and ensemble learning. The insights and findings from this work have implications for biomedical researchers, biocurators, and the text mining community working on biomedical relationship extraction.

Whole-exome sequencing identified first homozygous frameshift variant in the COLEC10 gene in an Iranian patient causing 3MC syndrome type 3

BACKGROUND

3MC syndrome type 3 is an autosomal recessive disorder caused by mutations in the COLEC10 gene besides other genes like COLEC11 and MASP1. This disorder is characterized by facial dysmorphism, cleft lip and palate, postnatal growth deficiency, cognitive impairment, hearing loss, craniosynostosis, radioulnar synostosis, genital and vesicorenal anomalies, cardiac anomalies, caudal appendage, and umbilical hernia.

METHODS

In the present study, whole-exome sequencing was performed in order to identify disease causing variant in an Iranian 7-year-old affected girl with craniosynostosis, dolichocephaly, blepharoptosis, clinodactyly of the 5th finger, high myopia, long face, micrognathia, patent ductus arteriosus, downslanted palpebral fissures, telecanthus, and epicanthus inversus. Identified variant confirmation in the patient and segregation analysis in her family were performed using Sanger sequencing method.

RESULTS

A novel homozygous frameshift deletion variant [NM_006438.5: c.128_129delCA; p.(Thr43AsnfsTer9)] was identified within the COLEC10 gene. Up to now, only three 3MC syndrome patients with mutations in the COLEC10 gene have been reported, and here, we report the fourth patient and the first homozygous frameshift variant.

CONCLUSION

Other genes and factors responsible for 3MC syndrome occurrence are remained to be discovered. We believe further investigation of the genes in the lectin complement pathway is needed to be done for the identification of other causes of this disease.

Dimensional reduction of phenotypes from 53 000 mouse models reveals a diverse landscape of gene function

Animal models have long been used to study gene function and the impact of genetic mutations on phenotype. Through the research efforts of thousands of research groups, systematic curation of published literature and high-throughput phenotyping screens, the collective body of knowledge for the mouse now covers the majority of protein-coding genes. We here collected data for over 53 000 mouse models with mutations in over 15 000 genomic markers and characterized by more than 254 000 annotations using more than 9000 distinct ontology terms. We investigated dimensional reduction and embedding techniques as means to facilitate access to this diverse and high-dimensional information. Our analyses provide the first visual maps of the landscape of mouse phenotypic diversity. We also summarize some of the difficulties in producing and interpreting embeddings of sparse phenotypic data. In particular, we show that data preprocessing, filtering and encoding have as much impact on the final embeddings as the process of dimensional reduction. Nonetheless, techniques developed in the context of dimensional reduction create opportunities for explorative analysis of this large pool of public data, including for searching for mouse models suited to study human diseases.

AVAILABILITY AND IMPLEMENTATION

Source code for analysis scripts is available on GitHub at https://github.com/tkonopka/mouse-embeddings. The data underlying this article are available in Zenodo at https://doi.org/10.5281/zenodo.4916171.

CONTACT

t.konopka@qmul.ac.uk.

SUPPLEMENTARY INFORMATION

Supplementary data are available at Bioinformatics Advances online.

A meta-analysis of prognostic biomarkers in neonatal retinal hemorrhage

PURPOSE

Neonatal retinal hemorrhage (RH) is a frequently occurring neonatal fundus condition and a very common ocular abnormality in neonates. Some of the key factors that influence the rate of RH are the mode of delivery, examination techniques, and time of examination after birth. The prognostic markers of severe RH are poorly known, making it difficult for an efficient diagnosis, prognosis, and treatment. Hence, to better understand the mechanism of disease, its study at the molecular level is required. Prognostic biomarkers are an essential tool for understanding the pathogenesis of the disease. In this paper, we present a meta-analysis of biomarkers to understand disease pathogenesis and support better diagnosis, prognosis, and treatment of neonatal RH.

METHODS

The meta-analysis was carried out by following the recommendation of PRISMA. The relevant articles were crawled using a systematic keyword using MeSH terms from the MEDLINE, PubMed, and Scopus databases, which were subjected to manual screening for reported biomarkers by two independent reviewers. The obtained biomarkers were further analyzed for gene-disease association and functional enrichment analysis.

RESULTS

Our meta-analysis suggests that genes ABCC6, Beta-APP, COL2A1, COL4A1, DNM2, ENPP1, IKBKG, ITGB2, IL-6, SELE, TREX1, and VEGFA are potential prognostic biomarkers associated with the neonatal RH. The gene-disease association and functional enrichment analysis suggest that few genes are associated with disease class "Vision"; however, some genes in the list are associated with the disease class "Pharmacogenomic," "Immune," "Renal."

CONCLUSION

The identified prognostic gene biomarkers may help to understand disease pathogenesis and provide a better diagnosis, prognosis, and treatment of neonatal RH.

Identification of a Rare Case With Nagashima-Type Palmoplantar Keratoderma and 18q Deletion Syndrome via Exome Sequencing and Low-Coverage Whole-Genome Sequencing

Nagashima-type palmoplantar keratoderma (NPPK) is characterized by non-progressive, diffuse, and cross-gradient hyperkeratosis caused by mutations in the SERPINB7 gene on chromosome 18q21.33. Chromosome 18q deletion syndrome (18q- syndrome) is a terminal deletion or microdeletion syndrome characterized by intellectual disability and congenital malformations. This paper describes an 18-year-old man with palmoplantar keratoderma and diffuse white matter abnormalities in the brain. Trio-based exome sequencing (ES) revealed a suspected mosaic compound heterozygous mutation for c.796C>T (p.Arg266^∗) in exon 8 inherited from the mother and a de novo exons 4–6 deletion of SERPINB7. Additional copy number variant (CNV) analysis of the ES data indicated a heterozygous gross deletion of 18q22.3-q23. The two SERPINB7 gene variants were verified by Sanger sequencing and quantitative real-time polymerase chain reaction (qRT-PCR). Finally, low-coverage whole-genome sequencing (WGS) confirmed the 18q22.3-q23 deletion and additionally detected a mosaic 18q21.33-q22.3 deletion, together explaining NPPK and the neurological phenotypes of the proband. The gross deletion of all exons of SERPINB7 was revealed for the first time. More rarely, c.796C>T (p.Arg266^∗) was likely to be mosaic, while the exon deletion was mosaic. In conclusion, the combination of multiple molecular genetic testing methods provides comprehensive informative molecular findings and promotes the diagnosis of complex diseases, as in this case.

Proposed criteria for nevoid basal cell carcinoma syndrome in children assessed using statistical optimization

Nevoid basal cell carcinoma syndrome (NBCCS) is a tumor predisposition condition, the cardinal features of which emerge in adolescence or adulthood. Using statistical optimization, this study proposes NBCCS criteria with improved sensitivity in children less than 18 years of age. Earlier detection may lead to improved surveillance and prevention of sequelae. A survey eliciting medical history was completed by, or on behalf of, individuals with NBCCS. Based on these findings, criteria for suspicion of NBCCS in children were suggested using information from a Bernoulli naïve Bayes classifier relying on the human phenotype ontology. The sensitivity and specificity of the existing and proposed diagnostic criteria were also assessed. Participants (n = 48) reported their first signs of NBCCS appeared at a median age of 8 months, but by our retrospective analysis, they did not fulfill the current diagnostic criteria until a median age of 7 years. This study delineates the early-onset features of NBCCS and proposes criteria that should prompt consideration of NBCCS. Additionally, we demonstrate a method for quantitatively assessing the utility of diagnostic criteria for genetic disorders.

Very Elevated IgE, Atopy, and Severe Infection: A Genomics-Based Diagnostic Approach to a Spectrum of Diseases

Elevated IgE has been long recognized as an important clinical marker of atopy but can be seen in a myriad of conditions. The discovery of autosomal dominant STAT3 deficiency marked the first recognition of hyper-IgE syndrome (HIES) and the first primary immunodeficiency linked to elevated IgE. Since then, genomic testing has increased the number of defects with associated mutations causing hyper-IgE syndrome and atopic diseases with FLG, DOCK8, SPINK5, and CARD11, among others. A spectrum of recurrent infections and atopy are hallmarks of elevated IgE with significant phenotypic overlap between each underlying condition. As treatment is predicated on early diagnosis, genomic testing is becoming a more commonly used diagnostic tool. We present a 6-year-old male patient with markedly elevated IgE and severe atopic dermatitis presenting with staphylococcal bacteremia found to have a heterozygous variant in FLG (p.S3247X) and multiple variants of unknown significance in BCL11B, ZAP70, LYST, and PTPRC. We review the genetic defects underpinning elevated IgE and highlight the spectrum of atopy and immunodeficiency seen in patients with underlying mutations. Although no one mutation is completely causative of the constellation of symptoms in this patient, we suggest the synergism of these variants is an impetus of disease.

Text Mining for Building Biomedical Networks Using Cancer as a Case Study

In the assembly of biological networks it is important to provide reliable interactions in an effort to have the most possible accurate representation of real-life systems. Commonly, the data used to build a network comes from diverse high-throughput essays, however most of the interaction data is available through scientific literature. This has become a challenge with the notable increase in scientific literature being published, as it is hard for human curators to track all recent discoveries without using efficient tools to help them identify these interactions in an automatic way. This can be surpassed by using text mining approaches which are capable of extracting knowledge from scientific documents. One of the most important tasks in text mining for biological network building is relation extraction, which identifies relations between the entities of interest. Many interaction databases already use text mining systems, and the development of these tools will lead to more reliable networks, as well as the possibility to personalize the networks by selecting the desired relations. This review will focus on different approaches of automatic information extraction from biomedical text that can be used to enhance existing networks or create new ones, such as deep learning state-of-the-art approaches, focusing on cancer disease as a case-study.

MitoPhen database: a human phenotype ontology-based approach to identify mitochondrial DNA diseases

Diagnosing mitochondrial disorders remains challenging. This is partly because the clinical phenotypes of patients overlap with those of other sporadic and inherited disorders. Although the widespread availability of genetic testing has increased the rate of diagnosis, the combination of phenotypic and genetic heterogeneity still makes it difficult to reach a timely molecular diagnosis with confidence. An objective, systematic method for describing the phenotypic spectra for each variant provides a potential solution to this problem. We curated the clinical phenotypes of 6688 published individuals with 89 pathogenic mitochondrial DNA (mtDNA) mutations, collating 26 348 human phenotype ontology (HPO) terms to establish the MitoPhen database. This enabled a hypothesis-free definition of mtDNA clinical syndromes, an overview of heteroplasmy-phenotype relationships, the identification of under-recognized phenotypes, and provides a publicly available reference dataset for objective clinical comparison with new patients using the HPO. Studying 77 patients with independently confirmed positive mtDNA diagnoses and 1083 confirmed rare disease cases with a non-mitochondrial nuclear genetic diagnosis, we show that HPO-based phenotype similarity scores can distinguish these two classes of rare disease patients with a false discovery rate <10% at a sensitivity of 80%. Enriching the MitoPhen database with more patients will improve predictions for increasingly rare variants.

The phenotypic spectrum of KCNT1: a new family with variable epilepsy syndromes including mild focal epilepsy

BACKGROUND AND OBJECTIVE

Pathogenic variants in KCNT1 have been associated with severe forms of epilepsy, typically sleep-related hypermotor epilepsy or epilepsy of infancy with migrating focal seizures. To show that pathogenic variants in KCNT1 can be associated with mild extra-frontal epilepsy, we report a KCNT1 family with a wide spectrum of phenotypes ranging from developmental and epileptic encephalopathy to mild focal epilepsy without cognitive regression and not consistent with sleep-related hypermotor epilepsy.

METHODS

A large Canadian family of Caucasian descent including 9 affected family members was recruited. Family members were phenotyped by direct interview and review of existing medical records. Clinical epilepsy gene panel analysis and exome sequencing were performed.

RESULTS

Phenotypic information was available for five family members of which two had developmental and epileptic encephalopathy and three had normal development and focal epilepsy with presumed extra-frontal onset. All three had predominantly nocturnal seizures that did not show hyperkinetic features. All three reported clusters of seizures at night with a feeling of being unable to breathe associated with gasping for air, choking and/or repetitive swallowing possibly suggesting insular or opercular involvement. Genetic analysis identified a heterozygous KCNT1 c.2882G > A, p.Arg961His variant that was predicted to be deleterious.

DISCUSSION

This family demonstrates that the phenotypic spectrum associated with KCNT1 pathogenic variants is broader than previously assumed. Our findings indicate that variants in KCNT1 can be associated with mild focal epilepsy and should not be excluded during variant interpretation in such patients based solely on gene-disease validity.

CADA: phenotype-driven gene prioritization based on a case-enriched knowledge graph

Many rare syndromes can be well described and delineated from other disorders by a combination of characteristic symptoms. These phenotypic features are best documented with terms of the Human Phenotype Ontology (HPO), which are increasingly used in electronic health records (EHRs), too. Many algorithms that perform HPO-based gene prioritization have also been developed; however, the performance of many such tools suffers from an over-representation of atypical cases in the medical literature. This is certainly the case if the algorithm cannot handle features that occur with reduced frequency in a disorder. With Cada, we built a knowledge graph based on both case annotations and disorder annotations. Using network representation learning, we achieve gene prioritization by link prediction. Our results suggest that Cada exhibits superior performance particularly for patients that present with the pathognomonic findings of a disease. Additionally, information about the frequency of occurrence of a feature can readily be incorporated, when available. Crucial in the design of our approach is the use of the growing amount of phenotype–genotype information that diagnostic labs deposit in databases such as ClinVar. By this means, Cada is an ideal reference tool for differential diagnostics in rare disorders that can also be updated regularly.

Exome Sequencing of 5 Families with Severe Early-Onset Periodontitis

Periodontitis is characterized by alveolar bone loss leading to tooth loss. A small proportion of patients develop severe periodontitis at the juvenile or adolescent age without exposure to the main risk factors of the disease. It is considered that these cases carry rare variants with large causal effects, but the specific variants are largely unknown. In this study, we performed exome sequencing of 5 families with children who developed stage IV, grade C, periodontitis between 3 and 18 y of age. In 1 family, we found compound heterozygous variants in the gene CTSC (p.R272H, p.G139R), 1 of which was previously identified in a family with prepubertal periodontitis. Subsequent targeted resequencing of the CTSC gene in 24 patients <25 y of age (stage IV, grade C) identified the known mutation p.I453V (odds ratio = 4.06, 95% CI = 1.6 to 10.3, P = 0.001), which was previously reported to increase the risk for adolescent periodontitis. An affected sibling of another family carried a homozygous deleterious mutation in the gene TUT7 (p.R560Q, CADD score >30 [Combined Annotation Dependent Depletion]), which is implicated in regulation of interleukin 6 expression. Two other affected siblings shared heterozygous deleterious mutations in the interacting genes PADI1 and FLG (both CADD = 36), which contribute to the integrity of the environment-tissue barrier interface. Additionally, we found predicted deleterious mutations in the periodontitis risk genes ABCA1, GLT6D1, and SIGLEC5. We conclude that the CTSC variants p.R272H and p.I453V have different expressivity and diagnostic relevance for prepubertal and adolescent periodontitis, respectively. We propose additional causal variants for early-onset periodontitis, which also locate within genes that carry known susceptibility variants for common forms. However, the genetic architecture of juvenile periodontitis is complex and differs among the affected siblings of the sequenced families.

Phenotypic diversity, disease progression, and pathogenicity of MVK missense variants in mevalonic aciduria

Mevalonic aciduria (MVA) and hyperimmunoglobulinemia D syndrome (MKD/HIDS) are disorders of cholesterol biosynthesis caused by variants in the MVK gene and characterized by increased urinary excretion of mevalonic acid. So far, 30 MVA patients have been reported, suffering from recurrent febrile crises and neurologic impairment. Here, we present an in-depth analysis of the phenotypic spectrum of MVA and provide an in-silico pathogenicity model analysis of MVK missense variants. The phenotypic spectrum of 11 MVA patients (age range 0-51 years) registered in the Unified European Registry for Inherited Metabolic Disorders database was systematically analyzed using terms of the Human Phenotype Ontology. Biochemical, radiological as well as genetic characteristics were investigated. Six of eleven patients have reached adulthood and four have reached adolescence. One of the adolescent patients died at the age of 16 years and one patient died shortly after birth. Symptoms started within the first year of life, including episodic fever, developmental delay, ataxia, and ocular involvement. We also describe a case with absence of symptoms despite massive excretion of mevalonic acid. Pathogenic variants causing MVA cluster within highly conserved regions, which are involved in mevalonate and ATP binding. The phenotype of adult and adolescent MVA patients is more heterogeneous than previously assumed. Outcome varies from an asymptomatic course to early death. MVK variants cluster in functionally important and highly conserved protein domains and show high concordance regarding their expected pathogenicity.

Solving patients with rare diseases through programmatic reanalysis of genome-phenome data

Reanalysis of inconclusive exome/genome sequencing data increases the diagnosis yield of patients with rare diseases. However, the cost and efforts required for reanalysis prevent its routine implementation in research and clinical environments. The Solve-RD project aims to reveal the molecular causes underlying undiagnosed rare diseases. One of the goals is to implement innovative approaches to reanalyse the exomes and genomes from thousands of well-studied undiagnosed cases. The raw genomic data is submitted to Solve-RD through the RD-Connect Genome-Phenome Analysis Platform (GPAP) together with standardised phenotypic and pedigree data. We have developed a programmatic workflow to reanalyse genome-phenome data. It uses the RD-Connect GPAP's Application Programming Interface (API) and relies on the big-data technologies upon which the system is built. We have applied the workflow to prioritise rare known pathogenic variants from 4411 undiagnosed cases. The queries returned an average of 1.45 variants per case, which first were evaluated in bulk by a panel of disease experts and afterwards specifically by the submitter of each case. A total of 120 index cases (21.2% of prioritised cases, 2.7% of all exome/genome-negative samples) have already been solved, with others being under investigation. The implementation of solutions as the one described here provide the technical framework to enable periodic case-level data re-evaluation in clinical settings, as recommended by the American College of Medical Genetics.

Donor genetic variants as risk factors for thrombosis after liver transplantation: A genome-wide association study

Thrombosis after liver transplantation substantially impairs graft- and patient survival. Inevitably, heritable disorders of coagulation originating in the donor liver are transmitted by transplantation. We hypothesized that genetic variants in donor thrombophilia genes are associated with increased risk of posttransplant thrombosis. We genotyped 775 donors for adult recipients and 310 donors for pediatric recipients transplanted between 1993 and 2018. We determined the association between known donor thrombophilia gene variants and recipient posttransplant thrombosis. In addition, we performed a genome-wide association study (GWAS) and meta-analyzed 1085 liver transplantations. In our donor cohort, known thrombosis risk loci were not associated with posttransplant thrombosis, suggesting that it is unnecessary to exclude liver donors based on thrombosis-susceptible polymorphisms. By performing a meta-GWAS from children and adults, we identified 280 variants in 55 loci at suggestive genetic significance threshold. Downstream prioritization strategies identified biologically plausible candidate genes, among which were AK4 (rs11208611-T, p = 4.22 × 10^-05 ) which encodes a protein that regulates cellular ATP levels and concurrent activation of AMPK and mTOR, and RGS5 (rs10917696-C, p = 2.62 × 10^-05 ) which is involved in vascular development. We provide evidence that common genetic variants in the donor, but not previously known thrombophilia-related variants, are associated with increased risk of thrombosis after liver transplantation.

CAMT-MPL: congenital amegakaryocytic thrombocytopenia caused by MPL mutations - heterogeneity of a monogenic disorder - a comprehensive analysis of 56 patients

Congenital amegakaryocytic thrombocytopenia caused by deleterious homozygous or compound heterozygous mutations in MPL (CAMT-MPL) is a rare inherited bone marrow failure syndrome presenting as an isolated thrombocytopenia at birth progressing to pancytopenia due to exhaustion of hematopoietic progenitors. The analysis of samples and clinical data from a large cohort of 56 patients with CAMT-MPL resulted in a detailed description of the clinical picture and reliable genotype-phenotype correlations for this rare disease. We extended the spectrum of CAMT causing MPL mutations regarding number (17 novel mutations) and impact. Clinical courses showed great variability with respect to the severity of thrombocytopenia, the development of pancytopenia and the consequences from bleedings. The most severe clinical problems were (i) intracranial bleedings pre- and perinatally and the resulting long-term consequences, and (ii) the development of aplastic anemia in the later course of the disease. An important and new finding was that thrombocytopenia was not detected at birth in a quarter of the patients. The rate of non-hematological abnormalities in CAMT-MPL was higher than described so far. Most of the anomalies were related to the head region (brain anomalies, ocular and orbital anomalies) and consequences of intracranial bleedings. The present study demonstrates a higher variability of clinical courses than described so far and has important implications on diagnosis and therapy. The diagnosis CAMT-MPL has to be considered even for those patients who are inconspicuous in the first months of life or show somatic anomalies typical for other inherited bone marrow failure syndromes.

A global map of associations between types of protein posttranslational modifications and human genetic diseases

There are >200 types of protein posttranslational modifications (PTMs) described in eukaryotes, each with unique proteome coverage and functions. We hypothesized that some genetic diseases may be caused by the removal of a specific type of PTMs by genomic variants and the consequent deregulation of particular functions. We collected >320,000 human PTMs representing 59 types and crossed them with >4M nonsynonymous DNA variants annotated with predicted pathogenicity and disease associations. We report >1.74M PTM-variant co-occurrences that an enrichment analysis distributed into 215 pairwise associations between 18 PTM types and 148 genetic diseases. Of them, 42% were not previously described. Removal of lysine acetylation exerts the most pronounced effect, and less studied PTM types such as S-glutathionylation or S-nitrosylation show relevance. Using pathogenicity predictions, we identified PTM sites that may produce particular diseases if prevented. Our results provide evidence of a substantial impact of PTM-specific removal on the pathogenesis of genetic diseases and phenotypes.

•

There is an enrichment of disease-associated nsSNVs preventing certain types of PTMs

•

We report 215 pairwise associations between 18 PTM types and 148 genetic diseases

•

The removal of lysine acetylation exerts the most pronounced effect

•

We report a set of PTM sites that may produce particular diseases if prevented

Utility of Gene Panels for the Diagnosis of Inborn Errors of Metabolism in a Metabolic Reference Center

Next-generation sequencing (NGS) technologies have been proposed as a first-line test for the diagnosis of inborn errors of metabolism (IEM), a group of genetically heterogeneous disorders with overlapping or nonspecific phenotypes. Over a 3-year period, we prospectively analyzed 311 pediatric patients with a suspected IEM using four targeted gene panels. The rate of positive diagnosis was 61.86% for intermediary metabolism defects, 32.84% for complex molecular defects, 19% for hypoglycemic/hyperglycemic events, and 17% for mitochondrial diseases, and a conclusive molecular diagnosis was established in 2-4 weeks. Forty-one patients for whom negative results were obtained with the mitochondrial diseases panel underwent subsequent analyses using the NeuroSeq panel, which groups all genes from the individual panels together with genes associated with neurological disorders (1870 genes in total). This achieved a diagnostic rate of 32%. We next evaluated the utility of a tool, Phenomizer, for differential diagnosis, and established a correlation between phenotype and molecular findings in 39.3% of patients. Finally, we evaluated the mutational architecture of the genes analyzed by determining z-scores, loss-of-function observed/expected upper bound fraction (LOEUF), and haploinsufficiency (HI) scores. In summary, targeted gene panels for specific groups of IEMs enabled rapid and effective diagnosis, which is critical for the therapeutic management of IEM patients.

PhenoDB, GeneMatcher and VariantMatcher, tools for analysis and sharing of sequence data

Background

With the advent of whole exome (ES) and genome sequencing (GS) as tools for disease gene discovery, rare variant filtering, prioritization and data sharing have become essential components of the search for disease genes and variants potentially contributing to disease phenotypes. The computational storage, data manipulation, and bioinformatic interpretation of thousands to millions of variants identified in ES and GS, respectively, is a challenging task. To aid in that endeavor, we constructed PhenoDB, GeneMatcher and VariantMatcher.

Results

PhenoDB is an accessible, freely available, web-based platform that allows users to store, share, analyze and interpret their patients’ phenotypes and variants from ES/GS data. GeneMatcher is accessible to all stakeholders as a web-based tool developed to connect individuals (researchers, clinicians, health care providers and patients) around the globe with interest in the same gene(s), variant(s) or phenotype(s). Finally, VariantMatcher was developed to enable public sharing of variant-level data and phenotypic information from individuals sequenced as part of multiple disease gene discovery projects. Here we provide updates on PhenoDB and GeneMatcher applications and implementation and introduce VariantMatcher.

Conclusion

Each of these tools has facilitated worldwide data sharing and data analysis and improved our ability to connect genes to phenotypic traits. Further development of these platforms will expand variant analysis, interpretation, novel disease-gene discovery and facilitate functional annotation of the human genome for clinical genomics implementation and the precision medicine initiative.

Supplementary Information

The online version contains supplementary material available at 10.1186/s13023-021-01916-z.

Patients with PWS and related syndromes display differentially methylated regions involved in neurodevelopmental and nutritional trajectory

Background

Prader–Willi syndrome is a rare genetic neurodevelopmental disorder caused by a paternal deficiency of maternally imprinted gene expression located in the chromosome 15q11–q13 region. Previous studies have demonstrated that several classes of neurodevelopmental disorders can be attributed to either over- or under-expression of specific genes that may lead to impairments in neuronal generation, differentiation, maturation and growth. Epigenetic changes that modify gene expression have been highlighted in these disorders. One recent study focused on epigenetic analysis and compared patients with PWS with patients with other imprinting disorders. No study, however, has yet focused on epigenetics in patients with PWS specifically by comparing the mutations associated with this syndrome.

Objective

This study investigated the epigenetic modifications in patients with PWS and patients with PWS-related disorders caused by inactivation of two genes of the PWS chromosomal region, SNORD116 and MAGEL2. Our approach also aimed to compare the epigenetic modifications in PWS and PWS-related disorders.

Methods

We compared genome-wide methylation analysis (GWAS) in seven blood samples from patients with PWS phenotype (five with deletions of the PWS locus, one with a microdeletion of SNORD116 and one with a frameshift mutation of MAGEL2 presenting with Schaaf–Yang syndrome), as well as two control patients. Controls were infants that had been studied for suspicion of genetic diseases that was not confirmed by the genetic analysis and the clinical follow-up.

Results

The analysis identified 29,234 differentially methylated cytosines, corresponding to 5,308 differentially methylated regions (DMRs), which matched with 2,280 genes. The DMRs in patients with PWS were associated with neurodevelopmental pathways, endocrine dysfunction and social and addictive processes consistent with the key features of the PWS phenotype. In addition, the separate analysis for the SNORD116 and MAGEL2 deletions revealed that the DMRs associated with the SNORD116 microdeletion were found in genes implicated in metabolic pathways and nervous system development, whereas MAGEL2 mutations mostly concerned genes involved in macromolecule biosynthesis.

Conclusion

The PWS is associated with epigenetic modifications with differences in SNORD116 and MAGEL2 mutations, which seem to be relevant to the different associated phenotypes.

Understanding microcephaly through the study of centrosome regulation in Drosophila neural stem cells

Microcephaly is a rare, yet devastating, neurodevelopmental condition caused by genetic or environmental insults, such as the Zika virus infection. Microcephaly manifests with a severely reduced head circumference. Among the known heritable microcephaly genes, a significant proportion are annotated with centrosome-related ontologies. Centrosomes are microtubule-organizing centers, and they play fundamental roles in the proliferation of the neuronal progenitors, the neural stem cells (NSCs), which undergo repeated rounds of asymmetric cell division to drive neurogenesis and brain development. Many of the genes, pathways, and developmental paradigms that dictate NSC development in humans are conserved in Drosophila melanogaster. As such, studies of Drosophila NSCs lend invaluable insights into centrosome function within NSCs and help inform the pathophysiology of human microcephaly. This mini-review will briefly survey causative links between deregulated centrosome functions and microcephaly with particular emphasis on insights learned from Drosophila NSCs.

Design of Generalized Search Interfaces for Health Informatics

In this paper, we investigate ontology-supported interfaces for health informatics search tasks involving large document sets. We begin by providing background on health informatics, machine learning, and ontologies. We review leading research on health informatics search tasks to help formulate high-level design criteria. We use these criteria to examine traditional design strategies for search interfaces. To demonstrate the utility of the criteria, we apply them to the design of ONTology-supported Search Interface (ONTSI), a demonstrative, prototype system. ONTSI allows users to plug-and-play document sets and expert-defined domain ontologies through a generalized search interface. ONTSI’s goal is to help align users’ common vocabulary with the domain-specific vocabulary of the plug-and-play document set. We describe the functioning and utility of ONTSI in health informatics search tasks through a workflow and a scenario. We conclude with a summary of ongoing evaluations, limitations, and future research.

ACO2 clinicobiological dataset with extensive phenotype ontology annotation

Pathogenic variants of the aconitase 2 gene (ACO2) are responsible for a broad clinical spectrum involving optic nerve degeneration, ranging from isolated optic neuropathy with recessive or dominant inheritance, to complex neurodegenerative syndromes with recessive transmission. We created the first public locus-specific database (LSDB) dedicated to ACO2 within the "Global Variome shared LOVD" using exclusively the Human Phenotype Ontology (HPO), a standard vocabulary for describing phenotypic abnormalities. All the variants and clinical cases listed in the literature were incorporated into the database, from which we produced a dataset. We followed a rational and comprehensive approach based on the HPO thesaurus, demonstrating that ACO2 patients should not be classified separately between isolated and syndromic cases. Our data highlight that certain syndromic patients do not have optic neuropathy and provide support for the classification of the recurrent pathogenic variants c.220C>G and c.336C>G as likely pathogenic. Overall, our data records demonstrate that the clinical spectrum of ACO2 should be considered as a continuum of symptoms and refines the classification of some common variants.

Genetic testing in dementia-A medical genetics perspective

OBJECTIVE

When a genetic cause is suspected in a person with dementia, it creates unique diagnostic and management challenges to the treating clinician. Many clinicians may be unaware of the practicalities surrounding genetic testing for their patients, such as when to test and what tests to use and how to counsel patients and their families. This review was conducted to provide guidance to clinicians caring for patients with dementia regarding clinically relevant genetics.

METHODS

We searched PubMed for studies that involved genetics of dementia up to March 2020. Patient file reviews were also conducted to create composite cases.

RESULTS

In addition to families where a strong Mendelian pattern of family history is seen, people with younger age of onset, especially before the age of 65 years were found to be at an increased risk of harbouring a genetic cause for their dementia. This review discusses some of the most common genetic syndromes, including Alzheimer disease, frontotemporal dementia, vascular dementia, Parkinson disease dementia/dementia with Lewy bodies and some rarer types of genetic dementias, along with illustrative clinical case studies. This is followed by a brief review of the current genetic technologies and a discussion on the unique genetic counselling issues in dementia.

CONCLUSIONS

Inclusion of genetic testing in the diagnostic pathway in some patients with dementia could potentially reduce the time taken to diagnose the cause of their dementia. Although a definite advantage as an addition to the diagnostic repository, genetic testing has many pros and cons which need to be carefully considered first.

Diffusion enables integration of heterogeneous data and user-driven learning in a desktop knowledge-base

Integrating reference datasets (e.g. from high-throughput experiments) with unstructured and manually-assembled information (e.g. notes or comments from individual researchers) has the potential to tailor bioinformatic analyses to specific needs and to lead to new insights. However, developing bespoke analysis pipelines from scratch is time-consuming, and general tools for exploring such heterogeneous data are not available. We argue that by treating all data as text, a knowledge-base can accommodate a range of bioinformatic data types and applications. We show that a database coupled to nearest-neighbor algorithms can address common tasks such as gene-set analysis as well as specific tasks such as ontology translation. We further show that a mathematical transformation motivated by diffusion can be effective for exploration across heterogeneous datasets. Diffusion enables the knowledge-base to begin with a sparse query, impute more features, and find matches that would otherwise remain hidden. This can be used, for example, to map multi-modal queries consisting of gene symbols and phenotypes to descriptions of diseases. Diffusion also enables user-driven learning: when the knowledge-base cannot provide satisfactory search results in the first instance, users can improve the results in real-time by adding domain-specific knowledge. User-driven learning has implications for data management, integration, and curation.

Development and application of the ocular immune-mediated inflammatory diseases ontology enhanced with synonyms from online patient support forum conversation

BACKGROUND

Unstructured text created by patients represents a rich, but relatively inaccessible resource for advancing patient-centred care. This study aimed to develop an ontology for ocular immune-mediated inflammatory diseases (OcIMIDo), as a tool to facilitate data extraction and analysis, illustrating its application to online patient support forum data.

METHODS

We developed OcIMIDo using clinical guidelines, domain expertise, and cross-references to classes from other biomedical ontologies. We developed an approach to add patient-preferred synonyms text-mined from oliviasvision.org online forum, using statistical ranking. We validated the approach with split-sampling and comparison to manual extraction. Using OcIMIDo, we then explored the frequency of OcIMIDo classes and synonyms, and their potential association with natural language sentiment expressed in each online forum post.

FINDINGS

OcIMIDo (version 1.2) includes 661 classes, describing anatomy, clinical phenotype, disease activity status, complications, investigations, interventions and functional impacts. It contains 1661 relationships and axioms, 2851 annotations, including 1131 database cross-references, and 187 patient-preferred synonyms. To illustrate OcIMIDo's potential applications, we explored 9031 forum posts, revealing frequent mention of different clinical phenotypes, treatments, and complications. Language sentiment analysis of each post was generally positive (median 0.12, IQR 0.01-0.24). In multivariable logistic regression, the odds of a post expressing negative sentiment were significantly associated with first posts as compared to replies (OR 3.3, 95% CI 2.8 to 3.9, p < 0.001).

CONCLUSION

We report the development and validation of a new ontology for inflammatory eye diseases, which includes patient-preferred synonyms, and can be used to explore unstructured patient or physician-reported text data, with many potential applications.

Pitfalls in Genetic Diagnostics: Why Phenotyping is Essential

New genetic testing technologies have revolutionized medicine within the past years. It is foreseeable that the development will continue with the introduction of new techniques. Nevertheless, despite improved technology, an exact clinical description of the phenotype is still necessary and it is important to critically question findings, both before initiating genetic testing and when interpreting the results. We present four brief case vignettes to point out difficulties associated with correctly interpreting genetic findings.

Towards realizing the vision of precision medicine: AI based prediction of clinical drug response

Accurate and individualized prediction of response to therapies is central to precision medicine. However, because of the generally complex and multifaceted nature of clinical drug response, realizing this vision is highly challenging, requiring integrating different data types from the same individual into one prediction model. We used the anti-epileptic drug brivaracetam as a case study and combine a hybrid data/knowledge-driven feature extraction with machine learning to systematically integrate clinical and genetic data from a clinical discovery dataset (n = 235 patients). We constructed a model that successfully predicts clinical drug response [area under the curve (AUC) = 0.76] and show that even with limited sample size, integrating high-dimensional genetics data with clinical data can inform drug response prediction. After further validation on data collected from an independently conducted clinical study (AUC = 0.75), we extensively explore our model to gain insights into the determinants of drug response, and identify various clinical and genetic characteristics predisposing to poor response. Finally, we assess the potential impact of our model on clinical trial design and demonstrate that, by enriching for probable responders, significant reductions in clinical study sizes may be achieved. To our knowledge, our model represents the first retrospectively validated machine learning model linking drug mechanism of action and the genetic, clinical and demographic background in epilepsy patients to clinical drug response. Hence, it provides a blueprint for how machine learning-based multimodal data integration can act as a driver in achieving the goals of precision medicine in fields such as neurology.

DTI-Voodoo: machine learning over interaction networks and ontology-based background knowledge predicts drug-target interactions

MOTIVATION

In silico drug-target interaction (DTI) prediction is important for drug discovery and drug repurposing. Approaches to predict DTIs can proceed indirectly, top-down, using phenotypic effects of drugs to identify potential drug targets, or they can be direct, bottom-up and use molecular information to directly predict binding affinities. Both approaches can be combined with information about interaction networks.

RESULTS

We developed DTI-Voodoo as a computational method that combines molecular features and ontology-encoded phenotypic effects of drugs with protein-protein interaction networks, and uses a graph convolutional neural network to predict DTIs. We demonstrate that drug effect features can exploit information in the interaction network whereas molecular features do not. DTI-Voodoo is designed to predict candidate drugs for a given protein; we use this formulation to show that common DTI datasets contain intrinsic biases with major effects on performance evaluation and comparison of DTI prediction methods. Using a modified evaluation scheme, we demonstrate that DTI-Voodoo improves significantly over state of the art DTI prediction methods.

AVAILABILITY

DTI-Voodoo source code and data necessary to reproduce results are freely available at https://github.com/THinnerichs/DTI-VOODOO.

SUPPLEMENTARY INFORMATION

Supplementary data are available at Bioinformatics online.

PhenoTagger: a hybrid method for phenotype concept recognition using human phenotype ontology

MOTIVATION

Automatic phenotype concept recognition from unstructured text remains a challenging task in biomedical text mining research. Previous works that address the task typically use dictionary-based matching methods, which can achieve high precision but suffer from lower recall. Recently, machine learning-based methods have been proposed to identify biomedical concepts, which can recognize more unseen concept synonyms by automatic feature learning. However, most methods require large corpora of manually annotated data for model training, which is difficult to obtain due to the high cost of human annotation.

RESULTS

In this article, we propose PhenoTagger, a hybrid method that combines both dictionary and machine learning-based methods to recognize Human Phenotype Ontology (HPO) concepts in unstructured biomedical text. We first use all concepts and synonyms in HPO to construct a dictionary, which is then used to automatically build a distantly supervised training dataset for machine learning. Next, a cutting-edge deep learning model is trained to classify each candidate phrase (n-gram from input sentence) into a corresponding concept label. Finally, the dictionary and machine learning-based prediction results are combined for improved performance. Our method is validated with two HPO corpora, and the results show that PhenoTagger compares favorably to previous methods. In addition, to demonstrate the generalizability of our method, we retrained PhenoTagger using the disease ontology MEDIC for disease concept recognition to investigate the effect of training on different ontologies. Experimental results on the NCBI disease corpus show that PhenoTagger without requiring manually annotated training data achieves competitive performance as compared with state-of-the-art supervised methods.

AVAILABILITYAND IMPLEMENTATION

The source code, API information and data for PhenoTagger are freely available at https://github.com/ncbi-nlp/PhenoTagger.

SUPPLEMENTARY INFORMATION

Supplementary data are available at Bioinformatics online.

Dynamic changes in RNA-protein interactions and RNA secondary structure in mammalian erythropoiesis

Two features of eukaryotic RNA molecules that regulate their post-transcriptional fates are RNA secondary structure and RNA-binding protein (RBP) interaction sites. However, a comprehensive global overview of the dynamic nature of these sequence features during erythropoiesis has never been obtained. Here, we use our ribonuclease-mediated structure and RBP-binding site mapping approach to reveal the global landscape of RNA secondary structure and RBP-RNA interaction sites and the dynamics of these features during this important developmental process. We identify dynamic patterns of RNA secondary structure and RBP binding throughout the process and determine a set of corresponding protein-bound sequence motifs along with their dynamic structural and RBP-binding contexts. Finally, using these dynamically bound sequences, we identify a number of RBPs that have known and putative key functions in post-transcriptional regulation during mammalian erythropoiesis. In total, this global analysis reveals new post-transcriptional regulators of mammalian blood cell development.

Disease ontologies for knowledge graphs

BACKGROUND

Data integration to build a biomedical knowledge graph is a challenging task. There are multiple disease ontologies used in data sources and publications, each having its hierarchy. A common task is to map between ontologies, find disease clusters and finally build a representation of the chosen disease area. There is a shortage of published resources and tools to facilitate interactive, efficient and flexible cross-referencing and analysis of multiple disease ontologies commonly found in data sources and research.

RESULTS

Our results are represented as a knowledge graph solution that uses disease ontology cross-references and facilitates switching between ontology hierarchies for data integration and other tasks.

CONCLUSIONS

Grakn core with pre-installed "Disease ontologies for knowledge graphs" facilitates the biomedical knowledge graph build and provides an elegant solution for the multiple disease ontologies problem.

Semantic similarity and machine learning with ontologies

Ontologies have long been employed in the life sciences to formally represent and reason over domain knowledge and they are employed in almost every major biological database. Recently, ontologies are increasingly being used to provide background knowledge in similarity-based analysis and machine learning models. The methods employed to combine ontologies and machine learning are still novel and actively being developed. We provide an overview over the methods that use ontologies to compute similarity and incorporate them in machine learning methods; in particular, we outline how semantic similarity measures and ontology embeddings can exploit the background knowledge in ontologies and how ontologies can provide constraints that improve machine learning models. The methods and experiments we describe are available as a set of executable notebooks, and we also provide a set of slides and additional resources at https://github.com/bio-ontology-research-group/machine-learning-with-ontologies.

A Biomedical Knowledge Graph System to Propose Mechanistic Hypotheses for Real-World Environmental Health Observations: Cohort Study and Informatics Application

BACKGROUND

Knowledge graphs are a common form of knowledge representation in biomedicine and many other fields. We developed an open biomedical knowledge graph-based system termed Reasoning Over Biomedical Objects linked in Knowledge Oriented Pathways (ROBOKOP). ROBOKOP consists of both a front-end user interface and a back-end knowledge graph. The ROBOKOP user interface allows users to posit questions and explore answer subgraphs. Users can also posit questions through direct Cypher query of the underlying knowledge graph, which currently contains roughly 6 million nodes or biomedical entities and 140 million edges or predicates describing the relationship between nodes, drawn from over 30 curated data sources.

OBJECTIVE

We aimed to apply ROBOKOP to survey data on workplace exposures and immune-mediated diseases from the Environmental Polymorphisms Registry (EPR) within the National Institute of Environmental Health Sciences.

METHODS

We analyzed EPR survey data and identified 45 associations between workplace chemical exposures and immune-mediated diseases, as self-reported by study participants (n= 4574), with 20 associations significant at P<.05 after false discovery rate correction. We then used ROBOKOP to (1) validate the associations by determining whether plausible connections exist within the ROBOKOP knowledge graph and (2) propose biological mechanisms that might explain them and serve as hypotheses for subsequent testing. We highlight the following three exemplar associations: carbon monoxide-multiple sclerosis, ammonia-asthma, and isopropanol-allergic disease.

RESULTS

ROBOKOP successfully returned answer sets for three queries that were posed in the context of the driving examples. The answer sets included potential intermediary genes, as well as supporting evidence that might explain the observed associations.

CONCLUSIONS

We demonstrate real-world application of ROBOKOP to generate mechanistic hypotheses for associations between workplace chemical exposures and immune-mediated diseases. We expect that ROBOKOP will find broad application across many biomedical fields and other scientific disciplines due to its generalizability, speed to discovery and generation of mechanistic hypotheses, and open nature.

Exploring the Impact of Cerebrovascular Disease and Major Depression on Non-diseased Human Tissue Transcriptomes

Background

The development of complex diseases is contributed by the combination of multiple factors and complicated interactions between them. Inflammation has recently been associated with many complex diseases and may cause long-term damage to the human body. In this study, we examined whether two types of complex disease, cerebrovascular disease (CVD) or major depression (MD), systematically altered the transcriptomes of non-diseased human tissues and whether inflammation is linked to identifiable molecular signatures, using post-mortem samples from the Genotype-Tissue Expression (GTEx) project.

Results

Following a series of differential expression analyses, dozens to hundreds of differentially expressed genes (DEGs) were identified in multiple tissues between subjects with and without a history of CVD or MD. DEGs from these disease-associated tissues-the visceral adipose, tibial artery, caudate, and spinal cord for CVD; and the hypothalamus, putamen, and spinal cord for MD-were further analyzed for functional enrichment. Many pathways associated with immunological events were enriched in the upregulated DEGs of the CVD-associated tissues, as were the neurological and metabolic pathways in DEGs of the MD-associated tissues. Eight gene-tissue pairs were found to overlap with those prioritized by our transcriptome-wide association studies, indicating a potential genetic effect on gene expression for circulating cytokine phenotypes.

Conclusion

Cerebrovascular disease and major depression cause detectable changes in the gene expression of non-diseased tissues, suggesting that a possible long-term impact of diseases, lifestyles and environmental factors may together contribute to the appearance of "transcriptomic scars" on the human body. Furthermore, inflammation is probably one of the systemic and long-lasting effects of cerebrovascular events.

Genetic testing in the diagnosis of chronic kidney disease: recommendations for clinical practice

The overall diagnostic yield of massively parallel sequencing–based tests in patients with chronic kidney disease (CKD) is 30% for paediatric cases and 6–30% for adult cases. These figures should encourage nephrologists to frequently use genetic testing as a diagnostic means for their patients. However, in reality, several barriers appear to hinder the implementation of massively parallel sequencing–based diagnostics in routine clinical practice. In this article we aim to support the nephrologist to overcome these barriers. After a detailed discussion of the general items that are important to genetic testing in nephrology, namely genetic testing modalities and their indications, clinical information needed for high-quality interpretation of genetic tests, the clinical benefit of genetic testing and genetic counselling, we describe each of these items more specifically for the different groups of genetic kidney diseases and for CKD of unknown origin.

Chromosomal Microarray Analysis Has a Poor Diagnostic Yield in Children with Developmental Delay/Intellectual Disability When Concurrent Cerebellar Anomalies Are Present

Chromosomal microarray analysis is commonly used as screening test for children with neurodevelopmental issues, also in case of complex neurological phenotypes. Developmental delay/intellectual disability is a common presentation sign in pediatric ataxias, diseases with high clinical and genetic heterogeneity. In order to determine the diagnostic yield of Array-CGH in such conditions, all the tests performed in the last 10-year activity of a single referral center in children who present, besides the neurodevelopmental impairment, cerebellar abnormalities have been systematically gathered. The study demonstrates that, except for Dandy-Walker malformation or poly-malformative phenotypes, chromosomal microarray analysis should be discouraged as first-line diagnostic test in pediatric ataxias with neurodevelopmental disability.

HEMDAG: a family of modular and scalable hierarchical ensemble methods to improve Gene Ontology term prediction

MOTIVATION

Automated protein function prediction is a complex multi-class, multi-label, structured classification problem in which protein functions are organized in a controlled vocabulary, according to the Gene Ontology (GO). "Hierarchy-unaware" classifiers, also known as "flat" methods, predict GO terms without exploiting the inherent structure of the ontology, potentially violating the True-Path-Rule (TPR) that governs the GO, while "hierarchy-aware" approaches, even if they obey the TPR, do not always show clear improvements with respect to flat methods, or do not scale well when applied to the full GO.

RESULTS

To overcome these limitations, we propose Hierarchical Ensemble Methods for Directed Acyclic Graphs (HEMDAG), a family of highly modular hierarchical ensembles of classifiers, able to build upon any flat method and to provide "TPR-safe" predictions, by leveraging a combination of isotonic regression and TPR learning strategies. Extensive experiments on synthetic and real data across several organisms firstly show that HEMDAG can be used as a general tool to improve the predictions of flat classifiers, and secondly that HEMDAG is competitive versus state-of-the-art hierarchy-aware learning methods proposed in the last CAFA international challenges.

AVAILABILITY

Fully-tested R code freely available at https://anaconda.org/bioconda/r-hemdag. Tutorial and documentation at https://hemdag.readthedocs.io.

SUPPLEMENTARY INFORMATION

Supplementary data are available at Bioinformatics online.

AnnotSV and knotAnnotSV: a web server for human structural variations annotations, ranking and analysis

With the dramatic increase of pangenomic analysis, Human geneticists have generated large amount of genomic data including millions of small variants (SNV/indel) but also thousands of structural variations (SV) mainly from next-generation sequencing and array-based techniques. While the identification of the complete SV repertoire of a patient is getting possible, the interpretation of each SV remains challenging. To help identifying human pathogenic SV, we have developed a web server dedicated to their annotation and ranking (AnnotSV) as well as their visualization and interpretation (knotAnnotSV) freely available at the following address: https://www.lbgi.fr/AnnotSV/. A large amount of annotations from >20 sources is integrated in our web server including among others genes, haploinsufficiency, triplosensitivity, regulatory elements, known pathogenic or benign genomic regions, phenotypic data. An ACMG/ClinGen compliant prioritization module allows the scoring and the ranking of SV into 5 SV classes from pathogenic to benign. Finally, the visualization interface displays the annotated SV in an interactive way including popups, search fields, filtering options, advanced colouring to highlight pathogenic SV and hyperlinks to the UCSC genome browser or other public databases. This web server is designed for diagnostic and research analysis by providing important resources to the user.

Strict network analysis of evolutionary conserved and brain-expressed genes reveals new putative candidates implicated in Intellectual Disability and in Global Development Delay

OBJECTIVES

Intellectual Disability (ID) and Global Development Delay (GDD) are frequent reasons for referral to genetic services and although they present overlapping phenotypes concerning cognitive, motor, language, or social skills, they are not exactly synonymous. Aiming to better understand independent or shared mechanisms related to these conditions and to identify new candidate genes, we performed a highly stringent protein-protein interaction network based on genes previously related to ID/GDD in the Human Phenotype Ontology portal.

METHODS

ID/GDD genes were searched for reliable interactions through STRING and clustering analysis was applied to detect biological complexes through the MCL algorithm. Six coding hub genes (TP53, CDC42, RAC1, GNB1, APP, and EP300) were recognised by the Cytoscape NetworkAnalyzer plugin, interacting with 1625 proteins not yet associated with ID or GDD. Genes encoding these proteins were explored by gene ontology, associated diseases, evolutionary conservation, and brain expression.

RESULTS

One hundred and seventy-two new putative genes playing a role in enriched processes/pathways previously related to ID and GDD were revealed, some of which were already postulated to be linked to ID/GDD in additional databases.

CONCLUSIONS

Our findings expanded the aetiological genetic landscape of ID/GDD and showed evidence that both conditions are closely related at the molecular and functional levels.

Inborn disorders of the malate aspartate shuttle

Over the last few years, various inborn disorders have been reported in the malate aspartate shuttle (MAS). The MAS consists of four metabolic enzymes and two transporters, one of them having two isoforms that are expressed in different tissues. Together they form a biochemical pathway that shuttles electrons from the cytosol into mitochondria, as the inner mitochondrial membrane is impermeable to the electron carrier NADH. By shuttling NADH across the mitochondrial membrane in the form of a reduced metabolite (malate), the MAS plays an important role in mitochondrial respiration. In addition, the MAS maintains the cytosolic NAD⁺ /NADH redox balance, by using redox reactions for the transfer of electrons. This explains why the MAS is also important in sustaining cytosolic redox-dependent metabolic pathways, such as glycolysis and serine biosynthesis. The current review provides insights into the clinical and biochemical characteristics of MAS deficiencies. To date, five out of seven potential MAS deficiencies have been reported. Most of them present with a clinical phenotype of infantile epileptic encephalopathy. Although not specific, biochemical characteristics include high lactate, high glycerol 3-phosphate, a disturbed redox balance, TCA abnormalities, high ammonia, and low serine, which may be helpful in reaching a diagnosis in patients with an infantile epileptic encephalopathy. Current implications for treatment include a ketogenic diet, as well as serine and vitamin B6 supplementation.

A system-level analysis of patient disease trajectories based on clinical, phenotypic and molecular similarities

MOTIVATION

Incorporating the temporal dimension into multimorbidity studies has shown to be crucial for achieving a better understanding of the disease associations. Furthermore, due to the multifactorial nature of human disease, exploring disease associations from different perspectives can provide a holistic view to support the study of their aetiology.

RESULTS

In this work, a temporal systems-medicine approach is proposed for identifying time-dependent multimorbidity patterns from patient disease trajectories, by integrating data from electronic health records with genetic and phenotypic information. Specifically, the disease trajectories are clustered using an unsupervised algorithm based on dynamic time warping and three disease similarity metrics: clinical, genetic and phenotypic. An evaluation method is also presented for quantitatively assessing, in the different disease spaces, both the cluster homogeneity and the respective similarities between the associated diseases within individual trajectories. The latter can facilitate exploring the origin(s) in the identified disease patterns. The proposed integrative methodology can be applied to any longitudinal cohort and disease of interest. In this article, prostate cancer is selected as a use case of medical interest to demonstrate, for the first time, the identification of temporal disease multimorbidities in different disease spaces.

AVAILABILITY AND IMPLEMENTATION

https://gitlab.com/agiannoula/diseasetrajectories.

SUPPLEMENTARY INFORMATION

Supplementary data are available at Bioinformatics online.